Regulatory T cells genetically modified for the lymphotoxin alpha gene and uses thereof

Abstract

The present invention relates to regulatory T cell and uses thereof. By their immunosuppressive and anti-inflammatory activities, regulatory T cells play a central role in peripheral tolerance and thus critically prevent the development of autoimmune and inflammatory disorders. The inventors showed that Foxp3+CD4+ Tregs express high levels of LTα, which negatively regulates their immunosuppressive signature. They demonstrated that the adoptive transfer of LTα−/− Tregs in mice protects from dextran sodium sulfate (DSS)-induced colitis and attenuates inflammatory bowel disease (IBD), multi-organ autoimmunity and the development of CAC. The inventors also showed that by mixed bone marrow chimeras that LTα expression specifically in hematopoietic cells negatively controls the immunosuppressive signature of Tregs. In particular, the present invention relates to regulatory T cell characterized in that it does not express or expresses reduced levels of lymphotoxin alpha.

Claims

1. A regulatory T cell that is genetically modified so that it does not express or expresses reduced levels of lymphotoxin alpha, wherein the regulatory T cell is genetically modified to silence a lymphotoxin alpha gene and wherein the regulatory T cell is also genetically modified to express a chimeric antigen receptor which recognizes and/or binds to an autoantigen.

2. The regulatory T cell of claim 1 wherein a gene coding for lymphotoxin alpha is deleted.

3. The regulatory T cell of claim 1 wherein a gene coding for lymphotoxin alpha is mutated resulting in a non-viable RNA.

4. A population of regulatory T cells according to claim 1.

5. A method of treating autoimmune disease in a subject in need thereof, comprising administering to the subject a therapeutically effective amount of the population of regulatory T cells according to claim 4.

6. The method according to claim 5 wherein the autoimmune disease is inflammatory bowel disease.

7. The method according to claim 5 wherein the autoimmune disease is multiple sclerosis or type 1 diabetes.

8. An ex vivo method for stimulating regulatory T cells immunosuppressive activity, said method comprising: i) obtaining a biological sample from a subject; ii) isolating regulatory T cells from said sample; iii) expanding isolated regulatory T cells in vitro; iv) genetically modifying said isolated regulatory T cells by silencing the lymphotoxin alpha gene; and v) transfecting or transducing the isolated regulatory T cell with a vector encoding a chimeric antigen receptor which recognizes and/or binds to an autoantigen.

9. The method according to claim 8 wherein the biological sample is a blood sample.

Description

FIGURES

(1) FIG. 1. Thymic Foxp3+ Tregs from LTα−/− mice show a highly suppressive signature. (A) The expression of Ltα and Ltβ was measured by qPCR in purified Foxp3-GFP-conventional CD4+SP thymocytes and Foxp3−GFP+CD4+CD8− Tregs from adult Foxp3-GFP reporter mice (n=3 experiments). (B) Representative histogram showing the expression of the cell surface LTα1β2 heterotrimer (detected by staining with the soluble LTβR-Fc protein) in conventional Foxp3− CD4+SP and Foxp3+CD4+ Treg cells from WT thymi (n=6). Data are pooled of 2 independent experiments (n=3-4 mice per group). (C) Representative histograms showing the expression of the cell surface LTα1β2 heterotrimer (detected by staining with the soluble LTβR-Fc protein) in Foxp3.sup.+ nTregs (n=20), Foxp3.sup.+ nTregs stimulated O/N with anti-CD3/CD28 antibodies (n=5), Foxp3.sup.+ iTregs (n=6) and CD8.sup.+CD28′° Tregs (n=6) derived from the spleen of WT mice. (D) The expression level of Il10, Ebi3, Tgfb1, Ifng, Gzmb and Fasl mRNAs was measured by qPCR in CD4+CD25+ thymic Tregs purified from WT (n=3-6) and LTα−/− (n=3-6) adult and postnatal d10 mice. Data are derived from 2 independent experiments (n=3-4 mice per group).

(2) FIG. 2. Peripheral LTα−/− Tregs exhibit an effector phenotype. The expression level of several genes known to be associated with the polarization and effector functions of Tregs was measured by qPCR in purified WT and LTα−/− splenic Tregs (n=3 mice per group).

(3) FIG. 3. The adoptive transfer of LTα−/− Tregs protects from the severity of DSS-induced colitis. (A) Representative flow cytometry profiles of Foxp3 expression in purified splenic CD4+CD25+ cells from WT and LTα−/− mice. (B) Experimental setup: colitis was induced by the administration of 2% DSS in drinking water for 7 days followed by water only until day 11 in WT mice injected 2 days before with 2×10.sup.5 WT or LTα−/− Tregs. Colon inflammation and CD4.sup.+ T cell priming in mesenteric lymph nodes were analyzed at day 11 and day 4 of the protocol, respectively. (C) Body weight loss relative to the initial weight on day 0 of WT mice injected with 2×10.sup.5 WT or LTα−/− Tregs. Data are derived from 3 independent experiment with 4 mice per group. (D) Disease activity index (DAI) was monitored during the course of DSS-induced colitis. (E) The histogram shows the histological score of the colon in both groups of mice. (F) The expression level of pro-inflammatory cytokines (Il6, Ifng, Tnfa, Il17a, Il1a and Il33) and chemokines (Ccl2 and Cxcl12) was measured by qPCR in colon tissues from mice injected with WT (n=4) or LTα−/− (n=4) Tregs at the end of the protocol. (G) Numbers of Ly6G+ neutrophils, F4/80+CD11b+ macrophages, CD11b+CD11c+ and CD8α+CD11c+ dendritic cells (DCs), CD19+B cells and CD4+ T cells observed in the colon of both groups. Data are derived from 2 independent experiments with 4 mice per group. (H) Flow cytometry profiles and numbers of Th1 (CD4+IFN-γ+) and Th17 (CD4+IL-17A+) colon-infiltrating T cells. Data are derived from 3 independent experiment with 4 mice per group. (I) Ratios of Treg/Th1 and Treg/Th17. (J) Body weight loss relative to the initial weight on day 0 of WT mice injected with 2.10.sup.5 WT or 1.10.sup.5 or 0.5.10.sup.5 LTα.sup.−/− Tregs. Data are derived from 2 to 3 independent experiments with 4 mice per group.

(4) FIG. 4. The adoptive transfer of LTα−/− Tregs treats from IBD. (A) Experimental setup: Rag2−/− recipient mice were adoptively transferred with CD4+CD25+ Treg-depleted CD45.1 WT splenocytes. Three weeks later when mice developed signs of IBD, they were injected with either 2×10.sup.5 WT or LTα−/− CD45.2 Tregs. Body weight was monitored during three weeks. (B) Body weight loss relative to the initial weight on day 0 of mice injected with WT or LTα−/− Tregs or untreated. (C) Histograms show colon length and the ratio of colon weight/length observed at the end of the protocol. (D) Flow cytometry profiles and numbers of total colon-infiltrating Foxp3+CD4+ Tregs. (E) Flow cytometry profiles and numbers of colon-infiltrating Foxp3+CD4+ Tregs of CD45.2 origin. (F) Representative flow cytometry profiles and numbers of Th1 (CD4+IFN-γ+) and Th17 (CD4+IL-17A+) colon-infiltrating cells of CD45.1 origin. (G) Ratios of Treg/Th1 and Treg/Th17 in colons.

(5) FIG. 5. The adoptive transfer of LTα−/− Tregs during colon chronic inflammation prevents the development of CAC in the AOM-DSS model. (A) Experimental setup: CD45.1 WT mice were injected with AOM, which initiates tumorigenesis, followed by three cycles of DSS in drinking water, inducing a chronic colitis, which promotes the development of colorectal tumors. 2×10.sup.5 splenic CD4+CD25+ Tregs purified either from CD45.2 WT or LTα−/− mice were adoptively-transferred in these mice before the first two cycles of DSS. Colons were collected at 3, 6 and 12 weeks after AOM injection. (B-C) Histograms show total numbers of tumors per colon (B) and their volumes (C) at 6 and 12 weeks of the protocol. (D) Histograms show the ratio of colon weight/length at 3, 6 and 12 weeks. (E) The expression level of pro-inflammatory cytokines was measured by qPCR in non tumoral colon tissues from mice injected with WT (n=4) or LTα−/− (n=4) Tregs at 3 and 6 weeks. (F) The histogram shows numbers of total colon infiltrating cells at 3 weeks. (G, H) Frequencies and numbers of Ly6G+ neutrophils, F4/80+CD11b+ and F4/80+CD11b− macrophages, CD11b+CD11c+ and CD11b−CD11c+ DCs, CD4+, CD8+ T cells, Th1, Th17 and CD4+Foxp3+ Tregs in the colon of both groups of mice. Data are derived from 2 independent experiment with 3-4 mice per group.

(6) FIG. 6. The adoptive transfer of Lta.sup.−/− Tregs attenuates the severity of multi-organ autoimmunity. (A) Experimental setup: Rag2.sup.−/− recipient mice were adoptively transferred with CD4.sup.+CD25.sup.+ Treg-depleted CD45.1 WT splenocytes. Four weeks later when mice start to loose body weight, they were injected with either 2×10.sup.5 WT or Lta.sup.−/− CD45.2 Tregs. Three weeks after Treg adoptive transfer, mice were sacrificed and peripheral tissues were examined for immune infiltrates. Rag2.sup.−/− recipients co-injected at the beginning of the protocol with CD4.sup.+CD25.sup.+ Treg-depleted CD45.1 WT splenocytes and CD45.2 WT Tregs were used as controls. (B) Body weight loss relative to the initial weight on day 0 of controls or mice transferred with WT or Lta.sup.−/− Tregs. (C) Diagrams representative of organ infiltration levels by CD45.1 donor cells normalized to the infiltration observed in controls. Each diagram represents one individual mouse.

(7) FIG. 7. The suppressive signature of Treg cells is controlled by the LTα1β2/LTβR axis. (A) Experimental setup: Lethally irradiated WT CD45.1×CD45.2 recipient mice were reconstituted with mixed BM cells from WT CD45.1+WT CD45.2 or WT CD45.1+LTα−/− CD45.2 (ratio 50:50). Six weeks later, CD45.2 WT and CD45.2 LTα−/− from WT and LTα−/− donor groups respectively were cell-sorted and analyzed for several genes associated with Treg suppressive functions. (B) Splenic CD4+ T cells of CD45.2 origin were analyzed by flow cytometry for the expression of Foxp3 in both groups. Histograms show frequencies and numbers of CD45.2 WT and LTα−/− CD4+Foxp3+ Tregs. (C) The expression level of Il10, Ebi3, Tgf-β1, Ifn-γ, Gzmb, Fasl and Il17a mRNAs was measured by qPCR in purified CD45.2 WT and LTα−/− Tregs from donor groups. (D) Purified WT Foxp3+CD4+ Tregs pre-incubated or not with a soluble LTβR-Fc fusion protein and co-cultured during 24 h with purified CD11c+ DCs were analyzed for the expression level of Klrg1, Il10, Ebi3, Tgfb1, Gzmb, Fasl and Il17a by qPCR.

(8) FIG. 8. LTα expression is conserved in human Tregs derived from peripheral blood. Expression of (A) intracellular LTα protein and (B) cell-surface LTα1β2 heterotrimer detected by staining with the soluble LTβR-Fc receptor was analyzed by flow cytometry in CD4.sup.+CD25.sup.−CD127.sup.lo Tregs derived from peripheral blood of male and female patients.

EXAMPLE

(9) Material & Methods

(10) Mice

(11) All mice—CD45.1 WT, CD45.1×CD45.2 WT, CD45.2 WT, CD45.2 LTα−/−, Rag2−/− and Foxp3-GFP reporter mice—were on a pure C57BL/6 background and maintained under specific pathogen free conditions at the CIML (France). Standard food and water were given ad libitum. Males and females were used at d10 after birth or at the age of 6-12 weeks. Chimeras were generated at 6-8 weeks of age. All procedures involving animals have been performed in accordance with the institutional and ethical guidelines.

(12) Healthy Volunteers Blood Collection and PBMCs Separation

(13) Blood was collected at the Etablissement Francais du Sang (Nantes, France) from healthy individuals. Written informed consent was provided according to institutional guidelines. PBMCs were isolated by FicollPaque density-gradient centrifugation (Eurobio, Courtaboeuf, France). Remaining red cells and platelets were eliminated with a hypotonic solution and centrifugation.

(14) BM Chimeras

(15) Before BM transplantation, mice were lethally irradiated with Cs-137 γ-radiation source (2 doses of 500 rads) and transplanted 24 h later with 107 BM cells from CD45.1 WT donor with CD45.2 donor (either WT or LTα−/− mice) at ratio 50:50. T-cell reconstitution was assessed by analyzing blood cells by flow cytometry. Mice were analyzed 6 weeks post-reconstitution.

(16) Treg Cell Isolation

(17) Thymic and splenic Treg cells were isolated by scratching thymus and spleen through a 70 μm mesh. Splenic red blood cells were lysed with lysis buffer (eBioscience). Before cell-sorting, CD4+ T cells were pre-enriched by depletion of CD8+ and CD19+ cells using anti-CD8α (clone 53.6.7) and anti-CD19 (clone 1D3) biotinylated antibodies with anti-biotin microbeads by AutoMACS (Miltenyi Biotech) via the deplete program. CD4+CD25+ Tregs were sorted using a FACSAriaIII cell sorter (BD).

(18) LTα−/− Treg stability in vivo

(19) 2.10.sup.5 CD4+CD25+ splenic Tregs purified from CD45.1 WT and CD45.2 LTα−/− mice were adoptively transferred intravenously into sub-lethally irradiated CD45.1×CD45.2 WT recipient mice (ratio 50:50). Seven days after transfer, CD45.1 WT and CD45.2 LTα−/− CD4+CD25+ splenic Tregs were purified with a FACSAriaIII cell sorter (BD).

(20) RNA-Seq Experiments

(21) CD4.sup.+CD25.sup.+ splenic Tregs were cell-sorted from WT and LTα.sup.−/− mice. Two biological replicates were prepared for each condition. Total RNA was extracted using the RNeasy Micro Kit (Qiagen) and treated with DNase I. RNA-seq libraries were prepared using the TruSeq Stranded mRNA kit (Illumina) and sequenced with the Illumina HiSeq 2000 machine to generate datasets of single-end 50 bp reads. The reads were mapped to the mouse reference genome (mm10) using TopHat2 (version 2.0.12), then counted using Cufflinks or Cuffdiff (version 2.2.1) and the mm10 genome GTF gene annotation file (https://support.illumina.com/sequencing/sequencing_software/igenome.html). In addition to read counting, Cuffdiff performs between-sample normalization and was used to calculate the differential gene expression and its statistical significance in LTα.sup.−/− vs WT Tregs. Expression levels generated by Cufflinks, as fragments per kilobase of transcript per million mapped reads (FPKM), were processed by the Matrix2png program to generate heat maps of gene expression levels which were normalized to a mean value of 0 and a variance of 1 across the samples. Identification of biological processes accounting for transcriptomic differences between LTα.sup.−/− and WT Tregs was performed with GSEA in calculating the enrichment in expression of every gene set defining a Gene Ontology (GO) biological process (c5.bp.v5.1) and in selecting the processes that are the most enriched. A number of permutations of 10,000 and a “classic” scoring scheme were used to compute the level of enrichment or Normalized Enrichment Score (NES) of a gene set. Null expression values were removed from the analysis. GO biological processes with NES reaching significance (P value<0.05 and FDR<0.25) were selected. Since different GO processes could be defined by gene sets sharing a certain degree of gene overlap, a network representing the GSEA selected processes and their connections depending on their gene set similarities was carried out with Cytoscape. Groups of related GO processes were determined using EnrichmentMap choosing an “Overlap Coefficient” over 0.7. A cluster analysis was performed using ClusterMaker and the implemented “MCL cluster” method. For each cluster of enriched and connected GO biological processes, the process with the most significant enrichment was selected and its NES considered.

(22) DSS-Induced Colitis Experiments

(23) Two days before the induction of colitis, WT recipient mice were injected i.v. with 2.10.sup.5 CD4+CD25+ splenic Tregs sorted from WT or LTα−/− mice or alternatively with 1.10.sup.5 splenic LTα−/− Tregs when mentioned. The induction of colitis was assessed by given 2% DSS (Alfa Aesar) in drinking water for 7 days, followed by only water until sacrifice at d11. Body weight, rectal bleeding and stool consistency were monitored every day after DSS administration and used to determine the DAI.

(24) IBD Experiments

(25) Rag2−/− recipient mice were injected i.v. with 5.10.sup.5 CD4+CD25+ Treg-depleted naive CD4+ T cells purified from CD45.1 WT mice. After 3-4 weeks, 2.10.sup.5 CD4+CD25+ splenic Tregs sorted from CD45.2 WT or LTα−/− mice were injected i.v. Body weight was monitored once per week during the course of IBD.

(26) Colitis-Associated Cancer (CAC) Experiments

(27) WT CD45.1 recipient mice were injected i.p. with Azoxymethane (AOM, 12.5 mg/kg, Sigma). After 5 days, 2.5% DSS (Alfa Aesar) was given in the drinking water over 5 days, followed by 16 days of tap water. This cycle was repeated twice (5 days of 2.5% DSS and 4 days of 2% DSS). 2.10.sup.5 CD4+CD25+ splenic WT Tregs or LTα−/− were injected i.v. in these mice before the first two cycles of DSS. Colons were collected at 3, 6 and 12 weeks after AOM administration.

(28) Multi-Organ Autoimmunity Experiments

(29) Rag2.sup.−/− recipient mice were injected i.v. with 3.10.sup.6 CD4.sup.+CD25.sup.+ Treg-depleted splenocytes purified from CD45.1 WT mice. Four weeks later, 2.10.sup.5 CD4.sup.+CD25.sup.+ splenic Tregs from CD45.2 WT or Lta.sup.−/− mice were adoptively transferred i.v. Controls concomitantly received 2.10.sup.5 CD4.sup.+CD25.sup.+ splenic Tregs from CD45.2 WT mice and 3.10.sup.6 CD4.sup.+CD25.sup.+ Treg-depleted splenocytes from CD45.1 WT mice at the beginning of the protocol. Body weight was monitored once per week during the course of the protocol.

(30) Isolation of Lamina Propria Mononuclear Cells from Colonic Tissue

(31) Colons were cut into 0.5 cm pieces, washed in HBSS with 2% FCS, then incubated twice in HBSS 2 mM EDTA at 37° C. under rotation (15 min then 30 min). Pieces were filtered on 70 μm cell strainer and incubated in culture medium (10% FCS, 1% Penicillin-Streptomycin and 1.5% HEPES in RPMI medium) with 1 mg/ml Collagenase VIII (Sigma) at 37° C. under rotation during 45 min. Cells were filtered and isolated by centrifugation with 40/100% Percoll (Sigma) gradient for 20 min at 2100 rpm at room temperature.

(32) In Vitro Co-Culture Assays, Treg Activation and iTreg Generation

(33) For co-culture assays, 2.10.sup.3 cell-sorted total CD11c.sup.hi DCs, CD11c.sup.hiPDCA-1.sup.lo, Sirpα.sup.+CD11c.sup.hi PDCA-1.sup.lo or CD11c.sup.int PDCA-1.sup.hi were co-cultured during 24 h at 37° C. with 10.sup.4 purified CD4.sup.+CD25.sup.+ Tregs that were or not pre-incubated during 1 h with a soluble LTβR-Fc recombinant protein (2 μg/ml; R&D systems). For Treg activation, 5.10.sup.4 cell-sorted CD4.sup.+CD25.sup.+ Tregs were cultured on plastic bound previously coated with anti-CD3.sub.ε antibody (5 μg/ml; clone 2C11) in a culture medium containing soluble anti-CD28 (1 μg/ml; clone 37.51) in the presence of IL-2 (200 U/ml, Immunotools) and TGF-β (0.2 ng/ml, eBioscience). iTregs were generated in vitro by culturing purified CD4.sup.+CD25.sup.− cells on plastic bound previously coated with anti-CD3.sub.ε antibody (5 μg/ml; clone 2C11) in a culture medium containing soluble anti-CD28 (1 μg/ml; clone 37.51) in the presence of IL-2 (200 U/ml, Immunotools) and TGF-β (20 ng/ml, eBioscience) for 4 days.

(34) Flow Cytometry

(35) Anti-CD4 (RM4.5), CD8α (53.6.7), CD45.1 (A20), CD45.2 (104), CD44 (IM7), CD25 (PC61), CD11b (M1/70), CD19 (1D3), CD62L (MEL-14) and IFN-γ (XMG1.2) antibodies were from BD. Anti-CD69 (H1.2F3), CCR7 (4B12), Qa-2 (695H1-9-9), IL-10 (JESS-16E3), F4/80 (6F12), CD11c (N418), IL-17A (TC11-18H10.1) and CCR6 (29-2L17) antibodies were from BioLegend. Anti-Ly6G (RB6-8C5), KLRG1 (2F1), Ki-67 (SolA15) and Foxp3 (FJK-16s) were from eBioscience. Anti-S1p1 (713412) was from RnD Systems. For intracellular staining of Foxp3, IL-10, IFN-γ, IL-17A and Ki-67, cells were fixed, permeabilized and stained with the Foxp3 staining kit according to the manufacturer's instructions (eBioscience). For detection of cytokines, cells were stimulated for 3 h at 37° C. with phorbol 12-myristate 13-acetate (PMA; 10 ng/mL; Sigma) and ionomycine (1 μg/mL; Sigma) in the presence of Brefeldin A (5 μg/mL; BD). For staining with LTβR-Fc, cells were incubated with LTβR-Fc (R&D systems) at 1 μg/106 cells for 45 min on ice. LTβR-Fc staining was visualized using an Alexa Fluor 488-conjugated donkey antihuman IgG F(ab′)2 fragment (Jackson ImmunoResearch). Human anti-CD4 (OKT4), CD25 (BC96), CD127 (A019D5) antibodies were purchased from BioLegend. Stained cells were analyzed with FACSCanto II (BD) and data were analyzed using FlowJo software.

(36) Quantitative RT-PCR

(37) Total RNA was isolated with TRIzol (Invitrogen) and cDNA was synthesized with random oligo dT primers and Superscript II reverse transcriptase (Invitrogen). qPCR was performed with SYBR Premix Ex Taq master mix (Takara) on a ABI 7500 fast real-time PCR system (Applied Biosystem). Results were normalized to actin mRNA.

(38) Immunofluorescence Staining

(39) Immuno fluorescence staining on thymic sections was performed as described previously by using Alexa Fluor 488-conjugated anti-Foxp3 (FJK-16s; eBioscience) and anti-K14 (AF64, Covance Research) revealed with Cy3-conjugated anti-rabbit (Invitrogen). Sections were counterstained with 1 μg/ml DAPI and mounted with Mowiol (Calbiochem). Images were acquired with a LSM 780 Leica SPSX confocal microscope and quantified with ImageJ software.

(40) Statistical Analysis

(41) Statistical significance was assessed with GraphPad Prism 6 software using unpaired Student's t test or Mann-Whitney test. The two-way Anova test with Bonferroni correction was used for the analysis of tumor growth, the loss of weight and DAI. *, P<0.05; **, P<0.01; ***, P<0.001, ****, P<0.0001. Normal distribution of the data was assessed using d'Agostino-Pearson omnibus normality test. Error bars represent mean±SEM.

(42) Results

(43) Developing LTα.sup.−/− Tregs Exhibit a Signature of Highly Suppressive Cells from their Emergence in the Thymus

(44) We and others have previously reported that LTα is upregulated in the thymus upon positive selection in single positive thymocytes and particularly in CD4.sup.+ thymocytes by high affinity interactions with medullary thymic epithelial cells. Since Foxp3.sup.+ Treg cells are selected by high affinity TCR interactions with thymic stromal cells, we analyzed the expression level of LTα mRNA in purified Tregs from the thymus of Foxp3-GFP reporter mice. Strikingly, we found that CD4.sup.+Foxp3.sup.+ Tregs express ˜5-fold more LTα mRNA than conventional CD4.sup.+Foxp3.sup.− T cells (FIG. 1A). Similarly to LTα, LTβ mRNA was also overexpressed in CD4.sup.+Foxp3.sup.− Tregs compared to conventional CD4.sup.+Foxp3.sup.− T cells. The staining with a soluble LTβR-Fc fusion protein revealed that LTα protein was substantially more expressed in CD4.sup.+Foxp3.sup.+ Tregs than in conventional CD4.sup.+Foxp3.sup.− T cells, as a membrane anchored LTα1β2 heterocomplex (FIG. 1B), which only binds to LTβR receptor. Natural Treg cell development is a multistage process that leads to the development of Foxp3.sup.+CD25.sup.+ Tregs from Foxp3.sup.−CD25.sup.+ and Foxp3.sup.+CD25.sup.− cell precursors. Interestingly, LTβR-Fc staining was substantially higher in Foxp3.sup.+CD25.sup.− precursors and Foxp3.sup.+CD25.sup.+ mature Treg cells than in Foxp3.sup.−CD25.sup.+ precursors, indicating that LTα1β2 expression correlates with that of Foxp3 (data not shown). LTα1β2 expression was conserved in natural Tregs derived from the spleen (FIG. 1C). This expression increased in activated Foxp3.sup.+ Tregs with anti-CD3.sub.ε/CD28 antibodies. We further examined whether LTα1β2 was expressed in other T-cell subsets endowed with regulatory properties such as induced Foxp3.sup.+ Tregs (iTregs) and CD8.sup.+CD28.sup.lo Tregs. In contrast to CD8.sup.+CD28.sup.lo Tregs, we found that the LTα1β2 heterocomplex was highly expressed in Foxp3.sup.+ iTregs (FIG. 1B). These data indicate that high levels of LTα1β2 are restricted to the Treg cell lineage expressing the transcription factor Foxp3 and that this expression substantially increases upon TCR activation. This preferential expression in Foxp3.sup.+ Tregs compared to conventional CD4.sup.+ T cells suggests that LTα could be involved in Treg suppressive activity. We thus analyzed the developmental and functional properties of Foxp3.sup.+ Tregs derived from LTα.sup.−/− mice.

(45) We observed that LTα.sup.−/− mice showed normal frequencies and numbers of Foxp3.sup.−CD25.sup.+ and Foxp3.sup.+CD25.sup.− precursors and Foxp3.sup.+CD25.sup.+ mature Tregs in their thymi (data not shown). Furthermore, similarly to their WT counterparts, Foxp3.sup.+CD25.sup.+ thymic Tregs from LTα.sup.−/− mice exhibited comparable level of the chemokine receptor CCR7 involved in cortico-medullary migration of single positive thymocytes and were thus preferentially located in the medulla at a normal density (data not shown). Compared to Qa-2.sup.−Foxp3.sup.+ newly generated Tregs, Qa-2.sup.+Foxp3.sup.+ mature Tregs from LTα.sup.−/− mice also upregulated the expression of the sphingo lipid receptor S1P1 (data not shown), implicated in T-cell egress from the thymus, suggesting that LTα.sup.−/− Tregs are normally exported to the periphery. In accordance with this observation, normal frequencies and numbers of recent thymic emigrants Treg cells were observed in the blood and spleen of LTα.sup.−/− mice (data not shown).

(46) We next investigated the expression level of several genes associated with Treg cell function by qPCR. In adult mice, although CTLA-4, CD39, CD73 and LAG-3 mRNAs showed normal expression levels (data not shown), in contrast, thymic LTα.sup.−/− Tregs expressed higher levels of IL-10, Ebi3, TGF-β1, IFN-γ, granzyme b (Gzmb) and FasL mRNAs compared to their WT counterparts (FIG. 1D). We hypothesized that the highly suppressive signature of LTα.sup.−/− Tregs could be acquired from the emergence of Treg cells. To verify this and exclude any potential peripheral effects due to Treg recirculation into the adult thymus, we analyzed LTα.sup.−/− Tregs during the perinatal period that corresponds to the initial appearance of Tregs in the thymus. Similarly to adult Tregs, we found that levels of several genes associated with Treg suppressive function such as IL-10, Ebi3, IFN-γ and FasL were increased in LTα.sup.−/− perinatal Tregs (FIG. 1D). To definitively rule out that this highly immunosuppressive signature could be associated with recirculating Tregs, the expression of the chemokine receptor CCR6 that distinguishes developing from recirculating Tregs was analyzed. Normal frequencies and numbers of CCR6.sup.− developing and CCR6.sup.+ recirculating Tregs were observed in LTα.sup.−/− mice (data not shown), indicating that these mice do not show a defect in Treg recirculation. Furthermore, the expression of several genes associated with Treg cell function was substantially increased in both purified CCR6.sup.− developing and CCR6.sup.+ recirculating LTα.sup.−/− Tregs compared to their respective WT counterparts (data not shown), indicating that LTα.sup.−/− Tregs show a highly immunosuppressive signature from their development in the thymus.

(47) Similarly to conventional CD4.sup.+ T cells, the maturation of CD4.sup.+Foxp3.sup.+ Tregs upon positive selection is characterized by loss of CD69 and the acquisition of Qa2. We found higher frequencies and numbers of CD69.sup.−Qa2.sup.+ mature cells in developing CCR6.sup.− Tregs in LTα.sup.−/− thymi compared to WT thymi (data not shown). Strikingly, in the CCR6.sup.− developing Treg population, we found that the expression of several genes associated with Treg cell function was specifically increased in CD69.sup.−Qa2.sup.+ mature Tregs from LTα.sup.−/− mice compared to WT mice (data not shown). Altogether, these data show that the expression of LTα negatively controls the suppressive signature of developing Tregs from the Qa-2.sup.+ stage.

(48) LTα.sup.−/− Tregs Adopt Specialized Differentiation Programs

(49) To gain insights into the suppressive activity of LTα.sup.−/− Tregs, we analyzed the molecular signature of LTα.sup.−/− splenic Tregs by high-throughput RNA-seq (data not shown). Genes showing a significant variation in gene expression between WT and LTα.sup.−/− Tregs (P value≤0.05) and a fold change difference≥2 and ≤0.5 were considered as up and downregulated, respectively. We identified a total of 306 upregulated and 113 downregulated genes in LTα.sup.−/− Tregs compared to WT Tregs (data not shown). To better characterize the set of genes modulated in LTα.sup.−/− Tregs, we performed a Gene Ontology (GO) analysis. Genes overexpressed in LTα.sup.−/− Tregs were associated with eight main biological processes (data not shown). The top GO term hit for the set of input genes was associated with cell cycle process and cell proliferation. A heatmap of genes implicated in these categories revealed that many key regulators of cell proliferation were upregulated in LTα.sup.−/− Tregs such as Uhrf1, implicated in the proliferation and maturation of colonic Tregs. Consistently with these RNA-seq data, we observed higher frequencies of proliferating Ki-67.sup.+ cells in Foxp3.sup.+ Tregs from the spleen of LTα.sup.−/− mice than in WT mice (data not shown). A heatmap of genes associated with transcription also identified key regulators of this process such as Ahr whose activation was found to induce suppressive Tregs that prevent T-cell induced colitis.

(50) Tregs can adopt specialized differentiation programs that are controlled by several transcription factors that have been associated with helper T cell differentiation. We found that Tbx21, Irf4, Rorc, Bcl6 and Pparγ transcription factors expressed by effector Tregs specialized in controlling Th1, Th2, Th17, CD4 follicular helper effector T cells and fat-resident T cells respectively were strongly upregulated in LTα.sup.−/− Tregs (data not shown). This upregulation was also observed in CCR6.sup.− developing and CCR6.sup.+ recirculating thymic LTα.sup.−/− Tregs (data not shown). Consistently with these observations, many genes reported to be associated with helper T cell polarization were also upregulated in LTα.sup.−/− Tregs. Moreover, the transcription factor Blimp-1 (Prdm1 gene) that represents a common signature for all effector Tregs as well as Klrg1 and Tigit that characterize terminally activated and/or differentiated effector Tregs were upregulated in LTα.sup.−/− Tregs. In accordance with these data, increased frequencies of KLRG1.sup.+ cells and CD69.sup.+CD44.sup.+ effector cells were observed in CD4.sup.+Foxp3.sup.+ Tregs from the spleen of LTα.sup.−/− mice (data not shown). Furthermore, we found that LTα.sup.−/− Tregs express higher levels of CD44, Helios and Nur77 activation markers by flow cytometry (data not shown).

(51) RNA-seq data also revealed that LTα.sup.−/− Tregs expressed abundant amounts of mRNAs encoding for several genes associated with immunosuppressive functions of Tregs such as Ebi3, Il10, Tgf-β and Gzmb. Consistently with these observations, in contrast to CD69.sup.−CD44.sup.+ Treg cells, several genes associated with Treg effector functions such as Il10, Ebi3, Tgfb, Ifng, Gzmb and Fasl were specifically upregulated in CD69.sup.+CD44.sup.+ effector Treg cells at high levels in LTα.sup.−/ mice (data not shown). Similarly to thymic Tregs (data not shown), the expression of other genes associated with Treg effector functions such as Ctla4, CD39, CD73 and Lag3 were unchanged in splenic LTα.sup.−/− Tregs (data not shown). Importantly, the expression of several candidate genes in the distinct categories identified by RNA-seq analysis was confirmed by qPCR on purified WT and LTα.sup.−/− Tregs (FIG. 2). Altogether, these data thus show that LTα.sup.−/− Tregs are polarized and thus exhibit an activated/effector phenotype.

(52) The Adoptive Transfer of LTα.sup.−/− Tregs Protects from Ulcerative Colitis

(53) Given that LTα.sup.−/− Tregs highly express several genes implicated in Treg suppressive functions (FIGS. 1D and 2), we next evaluated whether the adoptive transfer of LTα.sup.−/− Tregs shows therapeutic benefits to protect from dextran sodium sulfate (DSS)-induced colitis. 2.10.sup.5 CD4.sup.+CD25.sup.+ cells that predominantly contain Foxp3.sup.+ Tregs (FIG. 3A) purified from WT or LTα.sup.−/− mice were injected into WT recipient mice two days before the induction of colitis with 2% DSS (FIG. 3B). We observed that mice injected with LTα.sup.−/− Tregs lost significantly less weight than those injected with WT Tregs (FIG. 3C). Moreover, the disease activity index (DAI), which combines stool consistency, rectal bleeding and weight loss was substantially less important in these mice (FIG. 3D). In accordance with the weight loss and DAI, these mice displayed less damages of the colonic epithelium (data not shown) and a reduced colitis histological score at the end of the experiment (FIG. 3E). We also observed a reduced expression of pro-inflammatory cytokines such as Il 6, Ifnγ, Tnf-α, Il17A, Il1α and Il33 as well as of chemokines implicated in the recruitment of immune cells such as Ccl2 and Cxcl12 in colons of mice transferred with LTα.sup.−/− Tregs (FIG. 3F). We further analysed the nature of colon-infiltrating immune cells by flow cytometry. Numbers of neutrophils, macrophages, dendritic cells, B cells and CD4.sup.+ T cells were drastically reduced in mice transferred with LTα.sup.−/− Tregs compared to those transferred with WT Tregs (FIG. 3G). A reduced infiltration of CD3.sup.+ and B220.sup.+ cells was confirmed on histological colon sections (data not shown). Numbers of Th1 and Th17 effector CD4.sup.+ T cells were also reduced (FIG. 3H). Consequently, Treg/Th1 and Treg/Th17 ratios were increased in the colon of mice transferred with LTα.sup.−/− Tregs (FIG. 3I). We then assessed the potential of Lta.sup.−/− Tregs to protect against colitis by reducing the number of adoptively transferred cells from 2.10.sup.5 to 1.10.sup.5 and then to 0.5.10.sup.5 cells. We observed that 1.10.sup.5 Lta.sup.−/− Tregs still shows a better protection than 2.10.sup.5 WT Tregs characterized by reduced weight loss (FIG. 3J). Interestingly, 0.5.10.sup.5 Lta.sup.−/− Tregs show the same protective effect than 2.10.sup.5 WT Tregs, indicating that Lta.sup.−/− Tregs are ˜4 times more suppressive in vivo than their WT counterparts.

(54) We next further determined whether the adoptive transfer of LTα.sup.−/− Tregs inhibits CD4.sup.+ T cell priming in mesenteric lymph nodes five days after the administration of DSS. Of note, we found that mice injected with LTα.sup.−/− Tregs already showed longer colon length and reduced colonic weight/length ratio at this time point, indicative of attenuated colon inflammation (data not shown). Strikingly, numbers of Th1 and Th17 effector CD4.sup.+ T cells were substantially reduced in mesenteric lymph nodes of these mice (data not shown), indicating that LTα.sup.−/− Tregs inhibit the conversion of nave CD4.sup.+ T cells into effectors. Altogether, these data show that the adoptive transfer of LTα.sup.−/− Tregs protects from the development of ulcerative colitis by dampening colon inflammation and the priming of pathogenic CD4.sup.+ T cells in mesenteric lymph nodes.

(55) The Adoptive Transfer of LTα.sup.−/− Tregs Promotes the Recovery from Inflammatory Bowel Disease

(56) We next investigated whether the adoptive transfer of LTα.sup.−/− Tregs could show benefits to cure inflammatory bowel disease (IBD). To address this issue, IBD was induced by transfer of CD4.sup.+CD25.sup.+ Treg-depleted naïve CD4.sup.+ T cells from CD45.1 WT congenic mice into Rag2.sup.−/− recipient mice and the development of IBD was monitored by assessing weight loss. Around 3 to 4 weeks after T cell adoptive transfer, when mice developed clinical symptoms of IBD characterized by diarrhea and weight loss, they received purified WT or LTα.sup.−/− Tregs and body weight was monitored once per week (FIG. 4A). Mice that did not receive Tregs were used as controls. Mice transferred with WT Tregs gained more weight than mice that did not receive Tregs, indicating that the transfer of WT Tregs ameliorates IBD (FIG. 4B). Interestingly, mice that were transferred with LTα.sup.−/− Tregs gained more weight than mice that received WT Tregs. Strikingly, these mice showed a higher colon length with a reduced colonic weight/length ratio, indicative of attenuated colon inflammation (FIG. 4C). In accordance with these observations, these mice exhibited a reduced histological score (data not shown). Importantly, numbers of total Foxp3.sup.+ Tregs and Foxp3.sup.+ Tregs of CD45.2 origin were more elevated in the colon of mice transferred with LTα.sup.−/− Tregs than those injected with WT Tregs (FIG. 4D-E). Furthermore, numbers of Th1 and Th17 effector CD4.sup.+ T cells were reduced in the colon of these mice (FIG. 4F). Consequently, Treg/Th1 and Treg/Th17 ratios were increased in these mice (FIG. 4G). Of note, in this experimental setting we observed that frequencies and numbers of CD4.sup.+ T cells were reduced in several peripheral tissues such as salivary glands, pancreas and lung, indicating that LTα.sup.−/− Tregs also control tissue infiltration of autoreactive CD4.sup.+ T cells (data not shown). Furthermore, numbers of CD44.sup.hiCD62L.sup.hi central and CD44.sup.hiCD62L.sup.lo effector memory CD4.sup.+ T cells were specifically reduced in these peripheral tissues (data not shown). Altogether, these data demonstrate that the adoptive transfer of LTα.sup.−/− Tregs is able to treat IBD and controls tissue infiltration of autoreactive CD4.sup.+ T cells.

(57) The Adoptive Transfer of LTα.sup.−/− Tregs Attenuates the Development of CAC

(58) Given that LTα.sup.−/− Tregs protect from DSS-induced colitis (FIG. 3) and that colon chronic inflammation can result in the initiation of CAC, we next evaluated whether the adoptive transfer of LTα.sup.−/− Tregs protects from the emergence of CAC. For this, we used a classical CAC protocol that consists in the administration of azoxymethane (AOM), initiating tumorigenesis followed by three cycles of DSS, inducing a chronic colitis, which promotes the development of multiple colorectal tumors (FIG. 5A). These repetitive cycles of DSS mimic active and remission phases of colon inflammation observed in patients. LTα.sup.−/− Tregs were adoptively transferred before the two first cycles of DSS, period that corresponds to the CAC inflammation phase that precedes the development of colorectal tumors. Interestingly, mice that received LTα.sup.−/− Tregs showed fewer colorectal tumors, mainly located in the distal colon, with a globally reduced volume than mice transferred with WT Tregs at both 6 and 12 weeks of the CAC protocol (FIG. 5B-C and data not shown). We thus hypothesized that the development of colorectal tumors is prevented by reduced colon inflammation in mice transferred with LTα.sup.−/− Tregs. Consistently, these mice show a reduced colonic weight/length ratio from 3 weeks until the end of the protocol, indicative of an attenuated colon inflammation (FIG. 5D). At 3 weeks of the CAC protocol, we observed a reduced expression of pro-inflammatory cytokines such as Il1α, Il1β, Tnf-α and Il17A in the colon of mice injected with LTα.sup.−/− Tregs (FIG. 5E). Reduced expression of pro-inflammatory cytokines also persisted at 6 weeks of the CAC protocol. Furthermore, the expression of chemokines implicated in the recruitment of immune cells such as Ccl2, Ccl4, Cxcl10 and Cxcl12 were also reduced in colons of these mice (FIG. 5E). We next examined the nature of colon infiltrating inflammatory immune cells by flow cytometry at 3 weeks of the CAC protocol (FIG. 5F-H). Numbers of total colon infiltrating immune cells was substantially reduced in mice transferred with LTα.sup.−/− Tregs compared to those transferred with WT Tregs (FIG. 5F). Numbers of neutrophils, macrophages and dendritic cells were specifically reduced in the colon of these mice (FIG. 5G). Furthermore, while numbers of CD4.sup.+ and CD8.sup.+ T cells were similar in the colon of both groups, numbers of Th17 effector CD4.sup.+ T cells were specifically reduced in mice that received LTα.sup.−/− Tregs (FIG. 5H). In contrast, frequencies of colon-infiltrating Foxp3.sup.+ Tregs were increased in these mice although their numbers were similar to those observed in mice transferred with WT Tregs. Altogether, these data show that the adoptive transfer of LTα.sup.−/− Tregs during colon chronic inflammation attenuates the development of colorectal tumors by dampening colon inflammation.

(59) The Adoptive Transfer of Lta.sup.−/− Tregs Limits Multi-Organ Autoimmunity

(60) We next evaluated the ability of Lta.sup.−/− Tregs to limit multi-organ autoimmunity in a model of wasting disease. CD4.sup.+CD25.sup.+ Treg-depleted total splenocytes isolated from CD45.1 WT congenic mice were transferred into Rag2.sup.−/− recipients. Four weeks later, when mice lost weight they received purified WT or Lta.sup.−/− Tregs and body weight was monitored once per week (FIG. 6A). Mice that received concomitantly CD4.sup.+CD25.sup.+ Treg-depleted total splenocytes and WT Tregs at the beginning of the experimental protocol were used as controls. Interestingly mice transferred with Lta.sup.−/− Tregs gained more weight than mice that received WT Tregs (FIG. 6B). Importantly, these mice regained around 15% of their initial weight, almost reaching the weight of controls at the end of the protocol. Three weeks after Treg adoptive transfer, in contrast to mice that received WT Tregs, showing elevated numbers of splenic CD4.sup.+ and CD8.sup.+ T cells of CD45.1 origin, mice transferred with Lta.sup.−/− Tregs had similar numbers of these cells than those observed in controls (data not shown). Moreover, splenic CD4.sup.+ and CD8.sup.+ donor T cells contained less CD44.sup.+CD62L.sup.− activated cells in mice transferred with Lta.sup.−/− Tregs compared to mice injected with WT Tregs, indicating that Lta.sup.−/− Tregs attenuate T cell activation (data not shown). Similarly to controls, a reduced infiltration of inflammatory cells was observed by histological examinations in peripheral tissues, including the salivary glands and pancreas in mice transferred with Lta.sup.−/− Tregs compared to mice that received WT Tregs (data not shown). Accordingly, the examination of CD45.1 donor cell infiltration by flow cytometry in individual mice revealed a lower infiltration in the salivary glands, pancreas and kidney of mice transferred with Lta.sup.−/− Tregs (FIG. 6C). We took advantage of this setup based on the adoptive transfer of Treg-depleted total splenocytes (FIG. 6A) to assess the generation of autoantibodies against several peripheral organs. Immunostaining of Rag2.sup.−/− tissue sections with sera from the three groups of mice revealed that serum from mice transferred with Lta.sup.−/− Tregs contained less autoantibodies against salivary glands, pancreas, kidney, liver and lung than the serum of mice injected with WT Tregs (data not shown). Thus, the adoptive transfer of Lta.sup.−/− Tregs limits immune cell infiltrations and the generation of autoantibodies against several peripheral tissues.

(61) Adoptively Transferred LTα.sup.−/− Tregs Maintain their Highly Immunosuppressive Signature In Vivo

(62) Since Treg cells can show a certain plasticity, we analysed the stability of LTα.sup.−/− Tregs in vivo upon adoptive transfer. For this, sublethally irradiated CD45.1×CD45.2 WT recipient mice were transferred with the same ratio of cell-sorted CD4.sup.+CD25.sup.+ WT and LTα.sup.−/− Tregs of CD45.1 and CD45.2 origin, respectively (data not shown). One week after adoptive transfer, we purified CD4.sup.+CD25.sup.+ Tregs of both origins from the spleen of recipient mice (data not shown) and analysed the expression of several genes associated with Treg function. Similar frequencies and numbers of CD4.sup.+CD25.sup.+ cells of CD45.1 or CD45.2 origins were recovered (data not shown). However, LTα.sup.−/− Tregs of CD45.2 origin expressed high levels of Klrg1, Il10, Tgfb, Ifng, gzmb and IL17a (data not shown), indicating that LTα.sup.−/− Tregs retained their highly immunosuppressive signature upon adoptive transfer.

(63) LTα Expression in Hematopoietic Cells and LTα1β2/LTβR Axis Negatively Control the Suppressive Signature of Treg Cells

(64) Because LTα.sup.−/− mice show a disorganized thymic and splenic microenvironment, we first analysed the contribution of non-hematopoietic stromal cells in the highly immunosuppressive phenotype of LTα.sup.−/− Tregs. For this, we generated bone marrow (BM) chimeras in which lethally irradiated CD45.2 WT or LTα.sup.−/− recipient mice were reconstituted with WT BM cells from CD45.1 congenic mice (WT CD45.1: WT and WT CD45.1: LTα.sup.−/− mice, respectively). Six weeks after BM transplantation, CD4.sup.+CD25.sup.+ Treg cells of CD45.1 donor origin were cell-sorted from the spleen and analysed for the expression of several genes associated with Treg effector function (data not shown). Similar frequencies and numbers of Foxp3.sup.+ Tregs were observed in both groups of mice (data not shown). Furthermore, the expression of Klrg1, Tgfb, Gzmb, Fasl and IL17a was similar in both groups of mice, indicating that non-hematopoietic cells are not implicated in the highly suppressive signature of Tregs observed in LTα.sup.−/− mice (data not shown).

(65) We next determined the respective contribution of the hematopoietic compartment by generating mixed bone marrow chimaeras in which lethally irradiated CD45.1×CD45.2 WT recipient mice were reconstituted with BM cells (50:50) from WT CD45.1 and WT CD45.2 (WT donor group), or WT CD45.1 and LTα.sup.−/− CD45.2 (LTα.sup.−/− donor group) (FIG. 7A). Six weeks later, we found increased frequencies and numbers of CD4.sup.+Foxp3.sup.+ Tregs derived from LTα.sup.−/− CD45.2 BM cells compared to those derived from WT CD45.2 BM cells in the spleen (FIG. 7B). Strikingly, purified LTα.sup.−/− CD45.2 Tregs showed increased expression of Il10, Ebi3, Tgfb1, Ifng, Gzmb, Fasl and IL17a genes compared to WT CD45.2 Tregs (FIG. 7C). These data indicate that the expression of LTα in hematopoietic cells negatively controls the immunosuppressive signature of Treg cells.

(66) Since we observed that Tregs express LTα, as a membrane anchored LTα1β2 heterocomplex (FIG. 1), we assessed the contribution of LTα1β2/LTβR axis in controlling the suppressive signature of Tregs. In particular, we analyzed whether blocking LTα1β2/LTβR interactions between Tregs and dendritic cells impacts the suppressive signature of Treg cells. For this, purified WT CD4.sup.+CD25.sup.+ Tregs pre-incubated or not with a soluble LTβR-Fc fusion protein were co-cultured with purified CD11c.sup.+ dendritic cells. Interestingly, Tregs that were pre-incubated with LTβR-Fc upregulated the expression of several genes associated with Treg suppressive function such as Klrg1, Il10, Ebi3, Tel, Gzmb, Fasl and Il17a compared to un-pretreated Tregs (FIG. 7D). These data thus indicate that LTα1β2/LTβR interactions between Tregs and dendritic cells negatively regulate the suppressive signature of Tregs.

(67) LTα Expression is Conserved in Human Tregs Derived from Peripheral Blood

(68) We next assessed whether LTα expression is conserved in human Tregs derived from peripheral blood of female and male healthy donors. Foxp3.sup.+CD4.sup.+ Tregs were classically identified as CD4.sup.+CD25.sup.+CD127.sup.lo cells. Intracellular LTα protein (FIG. 8A) and the cell-surface LTα1β2 heterocomplex (FIG. 8B) were substantially detected by flow cytometry in Tregs of all donors analyzed, indicating that this expression is conserved in mice to human.

(69) Discussion

(70) Several studies have identified numerous molecules implicated in the positive regulation of Treg cell development and function. In contrast, few reports have described signals that negatively regulate Treg function. Here, by analyzing distinct T cell populations endowed with regulatory properties, we found that Foxp3.sup.+ Tregs substantially express Lta, as a membrane anchored LTα1β2 heterocomplex. LTα is expressed in Foxp3.sup.+ Treg cells, from their development in the thymus at the CD25.sup.−Foxp3.sup.+ precursor stage. This expression is conserved in peripheral CD25.sup.+Foxp3.sup.+ Tregs. Similarly to LTβR.sup.−/− mice, LTα.sup.−/− mice do not show any obvious defect in CD4.sup.+Foxp3.sup.+ Treg cell development in the thymus. However, the signature of genes associated with suppressive functions was greatly enhanced both in thymic and peripheral LTα.sup.−/− Tregs, indicating that LTα negatively regulates their immunosuppressive signature. Our data show that this phenotype is detectable from the development of Tregs in the thymus since this highly suppressive signature was observed from their emergence of this cell type at the perinatal period and in developing CCR6.sup.− Treg cells in the adult. This phenotype is thus likely not due to recirculating peripheral Tregs but is rather developmental. Furthermore, since the expression of LTα in the thymus correlates with that of Foxp3, which dictates Treg cell identity, this suggests that Tregs express key molecules that tightly control their activity to prevent the cell to over-react and thus over-suppress immune reactions.

(71) Both thymic and splenic LTα.sup.−/− Tregs do not show any obvious defect in the expression of CD39, CD73 and CD25 implicated in metabolic disruption and of CTLA-4 and LAG-3 implicated in the modulation of antigen presentation. In contrast, they show increased expression of IL-10, TGF-β and IL-35 immunosuppressive cytokines and granzyme B, IFN-γ and FasL involved in cytotoxicity-mediated suppression mechanisms. Although further investigations are required to define precisely their suppressive mode of actions, our results nevertheless indicate that LTα.sup.−/− Tregs likely mediate their potent suppressive activity through the secretion of inhibitory cytokines and the expression of molecules involved in cytolysis of target cells. RNA-seq data confirmed that LTα.sup.−/− Tregs show an activated/effector phenotype, characterized by augmented expression of Blimp-1, Klrg1 and Tigit markers, described to distinguish terminally activated and/or differentiated effector Tregs. Consistently with these observations, higher frequencies of CD44.sup.hiCD69.sup.+ and KLRG1.sup.+ effector Treg cells were also observed in the spleen of LTα.sup.−/− mice by flow cytometry. LTα.sup.−/− Tregs also express high levels of the Ikaros family transcription factor, Helios. Interestingly, CD4.sup.+Foxp3.sup.+ Helios.sup.+ Tregs have been shown to possess a highly suppressive function within the bulk of CD4.sup.+CD25.sup.+ Treg population. Furthermore, forced expression of Helios enhances the suppressive function of Tregs whereas Helios knock-down results in decreased of the suppressive function both in vitro and in vivo. This high expression of Helios thus comforts the notion that LTα.sup.−/− Tregs are highly immunosuppressive. Splenic LTα.sup.−/− Tregs also express higher level of Nur77, which is an immediate early gene upregulated by TCR stimulation, suggesting that they were recently activated by antigens. Furthermore, RNA-seq data revealed that LTα.sup.−/− Tregs show a signature of highly proliferative cells, which was confirmed by high frequencies of Ki-67.sup.+ proliferative Treg cells by flow cytometry. Altogether, these data show that LTα.sup.−/− Tregs possess an activated/effector phenotype.

(72) Although the stability of the Treg phenotype is a debated issue, we observed that adoptively transferred LTα.sup.−/− Tregs retained their highly immunosuppressive signature in vivo at least 7 days after transfer. The stability of LTα.sup.−/− Treg phenotype suggests that the transfer of these suppressive cells could show benefits in pathological conditions. Given that LTα.sup.−/− Tregs show a highly immunosuppressive signature, we have evaluated whether the adoptive transfer of LTα.sup.−/− Tregs displays superior therapeutic benefits than WT Tregs in protecting and treating inflammatory bowel disorders. Interestingly, the transfer of LTα.sup.−/− Tregs protects from DSS-induced colitis and treats from IBD more efficiently than WT Tregs. This was reflected by a reduced body weight loss, a higher colon length and a reduced histological score in mice transferred with LTα.sup.−/− Tregs compared to mice injected with WT Tregs. Furthermore, we observed that LTα.sup.−/− Tregs substantially reduce colon inflammation and the infiltration of inflammatory immune cells. In the DSS-induced colitis model, we found that the transfer of LTα.sup.−/− Tregs before the induction of colitis reduces the priming and/or expansion of Th1 and Th17 pathogenic cells in mesenteric lymph nodes. Importantly, the ratios Treg/Th1 and Treg/Th17 were increased in the colon in both the DSS-induced colitis and IBD models, suggesting that Tregs can also exert their suppressive effects locally in this tissue.

(73) By their ability to suppress colon inflammation, the adoptive transfer of LTα.sup.−/− Tregs also attenuates the development of CAC, which is known to be promoted by chronic inflammation. This was illustrated from colon carcinogenesis at ˜6 weeks by a ˜3-fold reduction in numbers of colorectal tumors that showed smaller volumes than tumors from mice injected with WT Tregs. Importantly, this protective effect persisted until the end of the CAC protocol i.e. at ˜12 weeks even if it was less pronounced. Compared to mice transferred with WT Tregs, this attenuation in CAC development in mice that received LTα.sup.−/− Tregs is explained by a reduced colon inflammation observable from 3 weeks of the CAC protocol. This was characterized by a reduced (i) colon weight/length ratio, (ii) expression of pro-inflammatory cytokines and (iii) chemokines implicated in the recruitment of inflammatory immune cells into the colon. Altogether, these data indicate that compared to their WT counterparts, LTα.sup.−/− Tregs show a higher capacity to treat colitis and protect from both colitis and CAC development. LTα.sup.−/− Tregs thus show an augmented anti-inflammatory/immunosuppressive function than WT Tregs. By decreasing the number of adoptively transferred cells, we were able to determine that Ltα.sup.−/− Tregs are ˜4 times more suppressive in vivo than their WT counterparts.

(74) Importantly, mixed bone marrow chimeras showed that the activated/effector phenotype of LTα.sup.−/− Tregs is due to the specific loss of LTα expression in hematopoietic cells and likely not in non-hematopoietic stromal cells. Furthermore, our data revealed that LTα1β2/LTβR interactions between Tregs and dendritic cells, particularly Sirpα.sup.+ cDCs and pDCs, negatively control the suppressive signature of Treg cells, suggesting that a direct cell contact with antigen-present cells regulates Treg suppressive activity.

(75) Since LTα, expressed as a membrane anchored LTα1β2 heterocomplex, is conserved in human Tregs, the adoptive transfer of LTα.sup.−/− Tregs is expected to find therapeutic applications to prevent and/or treat other inflammatory and autoimmune disorders. Furthermore, the transfer of these cells could also be beneficial to protect from the development of other inflammation-induced cancers such as pancreatic, lung or bladder carcinoma, induced by chronic pancreatitis, bronchitis and cystitis, respectively.

(76) In conclusion, our study identified that LTα expression in Tregs fine-tunes the suppressive capacity of this cell type. LTα could thus represent an interesting new therapeutic target to increase Treg activity, which is expected to find clinical applications in the field of Treg cell therapy by reducing the required cell number and by efficiently treating inflammatory and autoimmune disorders and preventing the development of inflammation-induced cancers.

REFERENCES

(77) Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.